Our long-term objective is to map the genomic sequence onto protein structure and function. To achieve this goal, understanding the details of protein folding process is essential. Although the concept of a protein energy landscape has been established, one of the key challenges confronting the biophysical community is to obtain the direct information on protein folding process in atomic detail. We propose to develop a general computational approach based on our novel roadmap-based method to understand this process. Our general roadmap-based approach will give relative folding rates, locate folding pathways, obligatory intermediate states, off-pathway intermediates, transition states, and verify the cooperativity between binding and folding. Our approach will utilize a roadmap (or a graph) to capture most important features of protein conformation space and energy landscape as proposed in Aim 1, in turn, rich thermodynamic and kinetic information will be extracted from the roadmap and further analyzed by graph-based tools as proposed in Aim 2. We have recently obtained promising results in predicting protein folding pathways using our novel graph- theoretical approach enhanced reaction-path algorithm, which is part of our roadmap-based approach. We expect our roadmap-based approach will yield a comprehensive picture of folding mechanism. The proposed applications in Aim 3 will focus on several small proteins, which will allow us to learn fundamental principles regarding the following aspects of protein folding mechanism: (a) unifying features in protein folding;(b) hidden intermediate;(c) "downhill" folding;(d) cooperativity between binding and folding. Information concerning folding process is not only indispensible in mapping the genomic sequence onto protein structure and function, but also important in amyloid diseases and other human diseases associated with intrinsically disordered proteins. A deeper understanding of protein folding process can ultimately lead to better computational models for drug design. PUBLIC HEALTH RELEVANCE: Understanding protein folding process in atomic detail is indispensible in mapping the genomic sequence onto protein structure and function. Protein folding/unfolding and misfolding are implicated in amyloid diseases and other human diseases associated with intrinsically disordered proteins. A deeper understanding of protein folding process can ultimately lead to better computational models for drug design.